Evaluation of a pharmacist-led digital hypertension management program in a tertiary hospital in Singapore

Abstract

Objective

To determine if a pharmacist-led digital enablement for medication optimization (DEMO) program was non-inferior to standard physician consultation in achieving a blood pressure (BP) target of <140/90mmHg within six months. Secondary outcomes included assessing BP changes and time to exit in DEMO group, system usability scale (SUS), and willingness to pay (WTP) for the DEMO program.

Methods

A prospective cohort study was conducted in a hospital outpatient setting (Mar 2023 -Oct 2024). The intervention group used a mobile application to transmit home BP readings to a clinical dashboard, with pharmacist-led teleconsultations every 2 - 4 weeks, during which antihypertensives were adjusted. Control group received standard physician consultations.

Results

Using 1:4 nearest neighbour propensity score matching, 145 control patients were matched to 56 DEMO participants. The probability of achieving BP <140/90mmHg within six months was 53% (95% CI 44-62) in DEMO vs 39% (95% CI 31-47) in controls. DEMO participants had a 36% higher relative probability of achieving the BP target (RD 0.14, 95% CI 0.01-0.27; P=0.041). Mean change in systolic and diastolic BP in DEMO group was -16.4mmHg (95% CI -21.9 to -10.9; P<0.001) and of -7.7mmHg (95% CI -10.5 to -4.9; P<0.001) respectively. Median time to exit DEMO was 46 days (IQR 20.5–74). Mean SUS was 82.2±13.7, indicating excellent usability. Mean WTP was SGD$13.2 per month (95% CI 0-32.0).

Conclusion

The pharmacist-led DEMO program was non-inferior to standard physician consultation in achieving a BP target of 140/90mmHg within six months, with high usability and modest WTP.

Keywords

hypertension pharmacist willingness-to-pay telehealth digital health telemedicine telemonitoring blood pressure mobile phone remote patient monitoring self monitoring

Introduction

Hypertension is a major risk factor for cardiovascular disease and a leading cause of premature morbidity and mortality worldwide.¹ In Singapore, the 2022 National Population Health Survey reported that over one in three adult residents has hypertension. ² Effective management is essential, yet healthcare systems face increasing strain from rising chronic disease burden, population aging, and workforce shortages.³ These pressures highlight the need for sustainable care models that optimize efficiency and accessibility.

Telemedicine and telemonitoring have emerged as potential strategies to address these challenges.^4,5 A systematic review found that home blood pressure (BP) monitoring with telehealth support can improve BP control and, in some studies, reduce mortality.⁶ Singapore’s physician-led Primary Tech-Enhanced Care (PTEC) program has demonstrated the feasibility and effectiveness of digital hypertension management in primary care.⁷

In Singapore, patients with hypertension in tertiary hospital outpatient clinics have been managed through conventional physician consultations. In recent years, collaborative physician–pharmacist models have expanded, with pharmacist-led clinics increasingly managing hypertension condition. Locally, these pharmacist-led clinics are offering advantages such as shorter waiting time, timely medication adjustments and cost-effectiveness.⁸ Despite these advantages, the conventional clinic-based approach remains constrained by limited manpower and resources. Its episodic, face-to-face format also restricts the ability to provide continuous management for an expanding population of patients with chronic hypertension. Telemonitoring models enable pharmacist to review trends longitudinally, identify out-of-range readings, and implement interventions asynchronously. International studies have reported that pharmacist-led telemonitoring can achieve significant improvements in BP outcomes through continuous home BP monitoring, integration of patient-generated data, and guideline-based medication adjustments.^9–11 However, these studies are conducted largely in the community and its applicability to a tertiary hospital, where patients typically present with greater clinical complexity due to their multiple co-morbidities, has not yet been established. Moreover, our tertiary hospital serves a multi ethnic population which potentially allow a better refection of real-world application of care delivery. This gap highlights the need to evaluate the effectiveness and feasibility of such models locally.

This study aimed to evaluate whether hypertension management using the pharmacist-led Digital Enablement for Medication Optimisation (DEMO) program was non-inferior to a control of standard physician consultation. Secondary objectives include assessing mean change in systolic and diastolic BP and the time to exit in DEMO group, patient reported digital platform and device’s usability measured by System Usability Scale (SUS), and willingness to pay (WTP) for the DEMO program.

Methods

Setting and study design

We conducted a prospective cohort study of patients with hypertension enrolled in the DEMO program at Changi General Hospital (CGH), Singapore, between April 2023 and June 2024. A retrospective cohort of patients who received standard physician consultations during the same period served as the control group. The study was approved by the SingHealth Centralised Institutional Review Board (CIRB ref: 2022/2390).

Participants

Eligibility criteria

Adults aged 21 years or older with a documented diagnosis of hypertension were eligible for inclusion. Additional criteria included ownership of a smartphone operating on iOS 13 or later or Android 9 or later, an active mobile data plan of at least 1 GB per month, and willingness to participate in the telemonitoring intervention. Patients were excluded if they experienced significant device related technical limitations or were unable to access the national secure digital identity platform (SingPass), which was required for authentication within the program.

Recruitment and enrolment

Eligible patients were introduced to the intervention by clinical staff and provided informed consent prior to participation. Enrolment occurred continuously throughout the study period.

Intervention group

Participants in the intervention arm (DEMO group) were provided with either a Bluetooth-enabled BP monitor (A&D® Medical, Japan; manufactured in China) or a non-Bluetooth BP monitor (Omron® Healthcare, Japan; manufactured in Vietnam). The Health Discovery+ mobile application was installed on participants’ smartphones. They were instructed to measure their BP at least once weekly using the assigned monitor. Readings from Bluetooth-enabled monitors were automatically transmitted to the secure, web-based Health Discovery+ portal, whereas readings from non-Bluetooth devices required manual entry by participants. All BP measurements were captured in a clinical dashboard and reviewed by pharmacists during the scheduled teleconsultation appointment. Teleconsultations were conducted every 2–4 weeks, during which antihypertensive therapy was titrated in accordance with the ACC/AHA hypertension guidelines,¹² under a collaborative prescribing agreement with the primary physician (Figure 1).

Figure 1.

Overview of standard care and DEMO program.

Control group

Control patients attended routine physician-led consultations in outpatient clinics, where antihypertensives were adjusted at the physician’s discretion (Figure 1). Office BP readings were used for both groups to ensure comparability of outcomes.

Allocation

The intervention was offered as a voluntary digital service; thus, allocation was non-randomised. Patients with greater digital literacy were more likely to opt in, introducing potential self-selection bias.

Follow-up

Participants from intervention group were monitored prospectively from the time of enrolment. Participants exited the DEMO program upon achieving target BP of <140/90mmHg or after six months of participation, whichever occurred first.

Withdrawal

Participants from intervention group could withdraw at any time without impact on usual care. Consistent with research governance standards, data collected before withdrawal were retained for analysis.

Data collection

Patient demographics and clinical characteristics for both the intervention and control groups were extracted from the CGH electronic medical records.

For the intervention group, office BP measurements were recorded at both enrolment and exit. At the end of the study, participants also completed an exit questionnaire, which included a SUS survey and a WTP assessment to evaluate the usability of the telehealth intervention and their financial valuation of the program.

For the control group, BP data was extracted from the CGH electronic records for analysis if at least two measurements were available – one at baseline and at least one follow-up measurement within 6 months.

Statistical analysis

Sample size calculation

The non-inferiority margin was set at -0.10 based on pharmacists’ clinical expertise. Assuming an expected proportion of 0.29 with appropriate SBP within 6 months in the control group and 0.56 in the intervention group based on pilot data, the minimum sample size required to assess non-inferiority of intervention at 0.80 power and 0.025 significance level in a study with a 1:1 allocation ratio is 26 patients per group.¹³ Assuming an expected proportion of 0.77 with appropriate DBP within 6 months in the control group and 0.89 in the intervention group based on pilot data, the minimum sample size required is 45 patients per group.¹³ To assess both SBP and DBP within 6 months, we would require at least 45 patients per group. Assuming a 15% dropout and/or missingness rate, we would have to recruit at least 53 patients per group.

Participant characteristics and matching

Descriptive statistics of patient demographic and clinical characteristics were reported as number and percentage for categorical data, mean ± standard deviation (SD) for normally distributed data, and median and interquartile range (IQR) for non-normally distributed data. To assess if there were differences in baseline characteristics between DEMO and control patients, the independent samples t-test or Mann-Whitney U-test was used for continuous variables, while the chi-squared test or Fisher’s exact test was used for categorical variables.

Propensity score matching was performed to ensure balanced distributions of potential confounders between the DEMO and control group. Propensity scores were estimated using a logistic regression of group (DEMO/control) on baseline age (continuous), gender (male/female), race (Chinese/Malay/Indian/others), baseline SBP (continuous), baseline DBP (continuous), baseline diabetes mellitus (yes/no), liver disease (yes/no), chronic kidney disease (yes/no), congestive heart failure (yes/no), myocardial infarction (yes/no), cerebrovascular accident, transient ischemic attack, or stroke (yes/no), hyperlipidemia (yes/no), medical specialty that managed patient (endocrine/cardiovascular/renal), and baseline BMI. A calliper width corresponding to 0.2 times the standard deviation of the propensity scores was selected.¹⁴ We employed 1:4 nearest neighbor matching, i.e. up to four nearest control patients per DEMO patient, within the propensity score caliper with replacement. Balance diagnostics were assessed using the standardized mean difference (% bias) and a plot of the standardized % bias across covariates.

Outcomes

Effect of DEMO (vs. Control) on BP

To determine the effect of DEMO (vs. control) on BP, linear mixed-effects models were used to model follow-up systolic and diastolic BP using an identity link, compound symmetric covariance structure, and robust error variance. Fixed effects were intervention (DEMO or control), time (month 1, 2, 3, 4, 5, or 6), and baseline BP, with random intercept for patient to account for repeated BP measurements over time. Time-adjusted, model predicted mean BP per group and model estimated between-group difference in means with 95% confidence intervals (CIs) were reported. Mixed-effects modified poisson regression with robust error variance was used to model the likelihood of achieving the BP target of <140/90mmHg using a log link, compound symmetric covariance structure. Fixed effects and random intercept were similar to the linear mixed model. The time-adjusted, model-predicted mean BP levels were calculated as marginal means averaged over the observed distribution of follow-up measurements to reflect the cohort’s actual follow-up experience. Similarly, the time-adjusted, model-predicted probabilities of achieving BP targets were calculated as marginal predicted probabilities averaged over the observed distribution of follow-up measurements. For both models, the marginal predictions were standardized over the overall baseline BP distribution of the entire study sample and not evaluated at group-specific mean baseline BP values. Time-adjusted, model predicted likelihood per group and model estimated between-group risk difference and risk ratio with 95% CIs were reported. The lower bound of the 95% CI of the adjusted between-group risk difference of achieving BP target was compared to the non-inferiority margin (-0.10) to assess non-inferiority. We performed a sensitivity analysis using personalized BP targets based on ACC/AHA hypertension guidelines.¹²

BP changes and time to exit in DEMO group

To assess the BP changes in the DEMO group, we used the Wilcoxon signed-rank test to measure if there were differences in the average BP level in the last week prior to exit from DEMO vs. baseline. We reported mean and median time to exit from DEMO.

Patient-reported usability measured by SUS

To assess the usability of the BP monitor and mobile application among all DEMO patients and by BP monitor type (Bluetooth/manual), we reported the mean ± SD total SUS score and the number and percent of patients in each SUS category (excellent [>80.3]; good [68–80.3]; poor [51–67]; awful [<51]). The SUS score was validated in Lewis JR et al.¹⁵

WTP for DEMO program

To estimate the WTP for the DEMO program, we calculated the mean WTP and 95% CI for DEMO patients randomized to the open-ended or dichotomous choice group. The debiased, epsilon method of estimation was based on the method in Hofstetter et al.,¹⁶ where the participant’s estimated WTP is equivalent to the participant’s reported WTP in open-ended format minus (mean reported WTP in open-ended format minus mean reported WTP in dichotomous choice format), plus ε_i, where ε_i is randomly drawn from a simulated normal distribution with zero mean and a standard deviation equivalent to that of the reported open-ended WTP values.

All analyses were conducted using Stata 18 (College Station, TX: StataCorp LLC) and R version 4.5.2 (R Core Team, 2025; R Foundation for Statistical Computing, Vienna, Austria). This study is reported according to the Consolidated Standards of Reporting Trial (CONSORT) guidelines.

Results

Participant characteristics and matching

A total of 59/60 (98.3%) DEMO and 1,443/2,566 (56.2%) control patients had at least one follow-up BP measurement within six months from index BP measurement. Of the 1,443 control patients, we identified 145 as suitable matches for 56 DEMO patients. Three DEMO patients did not have suitable matches and were excluded from the analysis.

Propensity score matching balance diagnostic was used to check whether matching successfully created comparable groups. Following propensity score matching, there were negligible differences between groups for the matched variables (Table 1). The standardized mean and median % bias before matching were 24.2% and 26.9%, respectively, compared to 3.0% and 1.8% after matching. The range of the standardized % bias after matching was between -5.4% and 7.8% for all matched variables except baseline liver disease (-11.3%). There was no difference in the mean baseline age between the matched control and DEMO group (59.9 vs. 59.7 years; P=0.941). We observed similar results for female gender (42.9% vs. 42.7%; P=0.987), Chinese (60.7% vs. 58.4%; P=0.812), Malay (17.9% vs. 19.0%; P=0.872), Indian (10.7% vs. 10.9%; P=0.980), mean baseline SBP (150.3 vs. 150.5mmHg; P=0.946), mean baseline DBP (86.0 vs. 86.4mmHg; P=0.882), diabetes mellitus (48.2% vs. 49.6%; P=0.889), liver disease (10.7% vs. 14.4%; P=0.557), chronic kidney disease (35.7% vs. 38.4%; P=0.772), congestive heart failure (19.6% vs. 16.7%; P=0.686), myocardial infarction (19.6^ vs. 19.0%; P=0.937), cerebrovascular accident, transient ischemic attack, or stroke (7.1% vs. 5.8%; P=0.776), hyperlipidemia (1.8% vs. 1.8%; P=1.000), the cardiovascular specialty (41.1% vs. 41.5%; P=0.962), the endocrine specialty (35.7% vs. 35.3%; P=0.961), and baseline BMI (28.5 vs. 28.5kg/m²; P=0.931).

Table 1.

Patient characteristics after matching.

Variable and sample	Mean (continuous variable) or % (categorical variable)		P Value
Variable and sample	DEMO program (n=56)	Control (n=145)	P Value
Baseline age in years (continuous)	59.696	59.908	0.941
Female gender	42.857	42.708	0.987
Race (Chinese)	60.714	58.482	0.812
Race (Malay)	17.857	19.048	0.872
Race (Indian)	10.714	10.863	0.980
Baseline SBP in mmHg	150.27	150.48	0.946
Baseline DBP in mmHg	86.036	86.414	0.882
Diabetes mellitus	48.214	49.554	0.889
Liver disease	10.714	14.435	0.557
Chronic kidney disease	35.714	38.393	0.772
Congestive heart failure	19.643	16.667	0.686
Myocardial infarction	19.643	19.048	0.937
Cerebrovascular accident, Transient ischemic attack, or stroke	7.143	5.804	0.776
Hyperlipidemia	1.786	1.786	1.000
Specialty (Cardiovascular)	41.071	41.518	0.962
Specialty (Endocrine)	35.714	35.268	0.961
Baseline Body mass index in kg/m² (continuous)	28.49	28.588	0.931

DBP=Diastolic blood pressure. DEMO=Digital enablement for medication optimization. SBP=Systolic blood pressure.

Effect of DEMO (vs. Control) on BP

The time-adjusted, linear mixed model predicted mean follow-up SBP in the matched DEMO and control groups were 139.3 and 145.5, respectively, while the mean follow-up DBP were 79.6 and 81.2, respectively (Table 2). Compared to propensity score matched patients in the control group, patients in the DEMO group had 6.3mmHg lower mean SBP within six months from baseline (B -6.25, 95% CI -10.62 to -1.88; P=0.005) but no difference in mean DBP within six months from baseline (B -1.59, 95% CI -4.37 to 1.19; P=0.262).

Table 2.

Effect of DEMO versus control on follow-up blood pressure.

Outcome and effect measure	Time-adjusted, model predicted mean or likelihood		Estimated difference in means or risk difference/ratio (95% CI), DEMO program vs. control	P Value
Outcome and effect measure	DEMO program (n=56)	Control (n=145)		P Value
SBP in mmHg	139.3 (136.3 to 142.2)	145.5 (142.3 to 148.7)	-6.25 (-10.62 to -1.88)	0.005
DBP in mmHg	79.6 (77.6 to 81.6)	81.2 (79.3 to 83.2)	-1.59 (-4.37 to 1.19)	0.262
Likelihood of achieving BP target of <140/90 mmHg	0.53 (0.44 to 0.62)	0.39 (0.31 to 0.47)	–	–
Risk difference	–	–	0.14 (0.01 to 0.27)	0.041
Risk ratio	–	–	1.36 (1.01 to 1.82)	0.042
Likelihood of achieving CV risk-stratified BP target	0.32 (0.22 to 0.43)	0.30 (0.23 to 0.38)	–	–
Risk difference	–	–	0.02 (-0.13 to 0.17)	0.784
Risk ratio	–	–	1.07 (0.67 to 1.70)	0.782

BP=Blood pressure. CI=Confidence interval. CV=Cardiovascular. DBP=Diastolic blood pressure. DEMO=Digital enablement for medication optimization. SBP=Systolic blood pressure. SD=Standard deviation.

The time-adjusted, mixed-effects modified Poisson regression model predicted probability of achieving the BP target of <140/90mmHg within six months from baseline was 53% (95% CI 44 to 62) in the DEMO group and 39% (95% CI 31 to 47) in the control group (Table 2). The lower bound of the 95% CI of the adjusted between-group risk difference in probability of achieving BP target was 0.01, which was above the non-inferiority margin of -0.10 (Figure 2). Patients in the DEMO group had a 14% higher absolute probability of achieving the BP target (RD 0.14, 95% CI 0.01 to 0.27; P=0.041) and a 36% higher relative probability of achieving the BP target (RR 1.36, 95% CI 1.01 to 1.82; P=0.042). In a sensitivity analysis using personalized BP targets based on ACC/AHA hypertension guidelines, the adjusted between-group risk difference was 0.02 (95% CI -0.13 to 0.17). While the lower bound of the 95% CI exceeded the non-inferiority margin, the risk difference estimate of 0.02 was not incongruent with a non-inferiority conclusion in the main analysis.

Figure 2.

Non-inferiority of the DEMO program to the control for a uniform versus personalized blood pressure target.

BP changes and time to exit in DEMO group

Eighty-seven percent (52/60) of DEMO patients had at least one BP measurement in the last week prior to exit from DEMO (Table 3). Among these 52 patients, the mean SBP in the last week prior to exit from DEMO and at baseline were 130.8±8.9mmHg and 147.2±21.6mmHg, respectively, corresponding to a mean difference of -16.4mmHg (95% CI -21.9 to -10.9; P<0.001; Table 3). The mean DBP in the last week prior to exit from DEMO and at baseline were 76.6±9.4mmHg and 84.3±11.9mmHg, respectively, corresponding to a mean difference of -7.7mmHg (95% CI -10.5 to -4.9; P<0.001; Table 3). Among all 60 DEMO patients, the mean time to exit from DEMO was 52.5±36.7 days, while the median time to exit was 46 days (IQR 20.5–74).

Table 3.

Blood pressure changes and time to exit in DEMO group.

Outcome and summary statistic	Baseline	Last week prior to exit from DEMO	Mean or median difference (95% CI)	P Value
SBP in mmHg (n=52)
Mean±SD	147.2±21.6	130.8±8.9	-16.4 (-21.9 to -10.8)	<0.001
Median (IQR)	143 (132.5–165)	131.2 (124.5–135.0)	-12.8 (-20.1 to -8.7)	–
DBP in mmHg (n=52)
Mean±SD	84.3±11.9	76.6±9.4	-7.7 (-10.5 to -4.9)	<0.001
Median (IQR)	83 (77–92)	78.1 (70.5–84.3)	-8.8 (-10.5 to -4.2)	–
Time to exit from DEMO in days (n=60)
Mean	–	–	52.5 (43.0 to 62.0)	–
Median	–	–	46 (31.7 to 56)	–

CI=Confidence interval. DBP=Diastolic blood pressure. DEMO=Digital enablement for medication optimization. IQR=Interquartile range. SBP=Systolic blood pressure. SD=Standard deviation.

Patient-reported usability measured by SUS

Of the 60 DEMO patients, two declined the SUS questionnaire and three were lost to follow up. Fifty-five (91.7%) had available SUS data. The mean SUS score was 82.2±13.7, indicating excellent usability (Table 4). Fifty-eight percent (32/55) of patients had an excellent SUS score, 29% (16/55) had a good score, and 13% (7/55) had a poor or awful score. Similar results were observed among DEMO patients who used the Bluetooth vs. manual BP monitor. Participants in the DEMO group graded positively in the satisfaction survey. They reported that they would likely use the BP meter and mobile application frequently, and did not find it unnecessarily complex. Most find it easy to use, did not think they would need technical support to use it. They indicated that the functions were well integrated, there was not too much inconsistency, and believed that most people would learn to use it very quickly. Additionally, they did not find it awkward to use, felt confident using it, and felt that little prior learning was needed to get started with the BP meter and mobile application.

Table 4.

Usability of blood pressure monitor and mobile application.

Subgroup and outcome	Mean±SD or no. (%)
All DEMO patients (n=55)
SUS score	82.2±13.7
SUS category
Excellent (>80.3)	32 (58.2)
Good (68–80.3)	16 (29.1)
Poor (51–67)	5 (9.1)
Awful (<51)	2 (3.6)
DEMO patients who used Bluetooth blood pressure monitor (n=38)
SUS score	82.2±14.9
SUS category
Excellent (>80.3)	22 (57.9)
Good (68–80.3)	11 (29.0)
Poor (51–67)	3 (7.9)
Awful (<51)	2 (5.3)
DEMO patients who used manual blood pressure monitor (n=17)
SUS score	82.4±11.1
SUS category
Excellent (>80.3)	10 (58.8)
Good (68–80.3)	5 (29.4)
Poor (51–67)	2 (11.8)
Awful (<51)	0 (0)

DEMO=Digital enablement for medication optimization. SD=Standard deviation. SUS=System usability scale.

WTP for DEMO program

Among the 60 DEMO participants, half were randomized to receive the open-ended or dichotomous choice WTP question format. Two patients in the open-ended group and three in the dichotomous choice group declined the questionnaire or were lost to follow up. An additional two patients in the open-ended group did not enter a valid value. Of the remaining 26 patients in the open-ended group, the mean WTP was SGD$20.80 (95% CI 6.9 to 34.7; Table 5). Of the 27 patients in the dichotomous choice group, the mean WTP was SGD$10.7 (95% CI 4.0 to 17.3; Table 5). The mean monthly WTP using the de-biased, epsilon method was SGD$13.2 (95% CI 0 to 32.0; Table 5). An estimated 27% of DEMO patients would be willing to pay SGD$36 per month for the DEMO program.

Table 5.

Estimated willingness to pay for DEMO program.

Method of estimation	Willingness to pay in SGD$
Method of estimation	<36, no. (%)	≥36, no. (%)	Mean (95% CI)
Open-ended (n=26)	22/26 (84.6)	4/26 (15.4)	20.77 (6.87 to 34.66)
Dichotomous choice (n=27)	19/27 (70.4)	8/27 (29.6)	10.67 (4.04 to 17.29)
De-biased, epsilon method (n=26)	19/26 (73.1)	7/26 (26.9)	13.19 (0^† to 32.04)

†Negative value was converted to a lower bound of zero.

CI=Confidence interval. DEMO=Digital enablement for medication optimization.

No adverse events related to telemonitoring or pharmacist-led medication adjustments were reported during the study period.

Discussion

This prospective cohort study has shown that the pharmacist-led DEMO program, which utilized a mobile application and home BP monitor for telemonitoring, was non-inferior to standard physician consultations in achieving BP target of <140/90mmHg within 6 months. Our findings are consistent with prior studies reporting that telemonitoring achieved greater BP reductions than standard care⁷ and was effective when used in pharmacist-led primary care clinics internationally.^9,17 Previous studies have suggested that telemonitoring improved outcomes when combined with proactive support such as counselling, behaviour management and education.¹⁸ However, in our study, participants achieved BP control with home monitoring and pharmacist-led medication adjustments according to collaborative prescribing agreement, without requiring additional physician reviews. This may be attributed to the consistent and reliable home BP readings¹⁹ generated through the DEMO program, along with the shorter intervals between clinic visits made possible with pharmacist-led consultations,²⁰ compared to traditional physician clinic sessions. These factors likely contributed to more timely and safe medication titrations. In our study, pharmacist-led consultations were scheduled every 2 - 4 weeks, aligning with the recommendations by Sherman et al., who have suggested that a biweekly follow up as optimal frequency for effective hypertension management until BP is controlled.²¹ Additionally, pharmacist-led interventions have been associated with improved patient adherence to medications,^22,23 further supporting patients to achieve BP goals more quickly. These factors may account for the 36% higher relative probability of achieving BP targets in the DEMO intervention group compared with the control group.

This study demonstrated significant reductions in systolic and diastolic BP from baseline to the last week prior to the exit from the DEMO program (mean systolic BP change: –16.4 mmHg; 95% CI –21.9 to –10.9; mean diastolic BP change: –7.7 mmHg; 95% CI –10.5 to –4.9; P<0.001). While a reduction in BP from baseline was expected with any intervention, this study also demonstrated that the DEMO program achieved substantially shorter time to BP control. The mean and median time to exit from DEMO was 52 days and 46 days respectively. This is shorter than the 3.25 months reported in a real-world study required to achieve BP control, determined based on the availability of at least one follow up BP reading within 12 months of the index date.²⁴ In DEMO, participants underwent teleconsultation at least every four weeks until target BP was achieved, or up to six months, facilitating earlier detection and titration of therapy, and likely accounting for the faster attainment of BP control.

The usability of the mobile application and BP monitor was rated excellent, with a mean SUS score of 82.2 ± 13.7. Despite prior local study showing patient preference for Bluetooth-enabled devices,²⁵ SUS scores did not significantly differ between the Bluetooth and manual BP monitor groups. Future research could further explore usability from the clinician perspective.

The mean patients’ out-of-pocket WTP for the pharmacist-led DEMO program was at SGD$13.20 per month. It showed an encouraging signal that patients were open to modest out-of-pocket cost for such programs. Interestingly, our estimated WTP of SGD$13.20 per month was closely aligned with the Health Discovery+ mobile application subscription fee (SGD$ 13.22 per month), indicating potential financial sustainability of such programs. These findings, though limited by small sample size, may inform future pricing strategies to support broader adoption of telehealth services and enhance their accessibility and cost-effectiveness.

This study has several limitations. First, the non-randomized design and retrospective control group introduce potential selection bias, including differences in digital literacy and education level. Although robust adjustments were made to minimize confounding, participant assignment was not randomized, propensity score matching mitigated but did not eliminate residual confounding. Second, the study applied a uniform BP target of <140/90 mmHg rather than risk-stratified goals (<130/80 mmHg for high-risk groups, such as those with cardiovascular disease, chronic kidney disease, or diabetes, which may limit applicability.²⁶ Findings should therefore not be extrapolated to long-term BP maintenance or broader clinical outcomes, such as cardiovascular event reduction. Third, the limited sample size, constrained by funding, may also restrict generalizability, particularly for WTP analysis. Fourth, there were differences in BP monitoring frequency and clinical follow-up intervals between groups, hence it required the use of a time-adjusted, mixed-effects modified Poisson regression model to predict the probability of achieving the BP target of <140/90mmHg within six months from baseline. Finally, adverse effects, including orthostatic hypotension, were not systematically captured.

Despite the growing adoption of digital hypertension management program, several implementation challenges may limit the scalability and long-term sustainability. Substantial financial barriers such as ongoing platform development, maintenance costs, and unclear reimbursement models, remain particularly problematic in resource-constrained settings. Successful implementation also depends on robust digital infrastructure, and interoperability with existing electronic health record systems. The management of large volumes of patient-generated health data further introduces concerns regarding data privacy, cybersecurity and regulatory compliance. While pharmacist-led care models may alleviate some of the barriers by supporting structured medication management, guideline-based titration, and patient education, persistent gaps in digital literacy, especially among older adults or individuals with low health literacy, pose added challenges to sustained engagement and long-term adherence. Collectively, these factors underscore the need for comprehensive implementation strategies that address both technological and human-centered barriers.

Conclusion

The pharmacist-led DEMO program was found to be non-inferior to standard physician consultations in achieving the target BP of 140/90 mmHg within six months. Users reported excellent usability of both the mobile application and the BP monitoring device, and the estimated willingness to pay was SGD$13.20 per month.

These findings suggested feasibility and potential scalability of pharmacist-led DEMO program; however, achieving broader scale-up and long-term sustainability of digital program will require comprehensive and well-coordinated implementation strategies.

Supplemental material

Supplemental material - Evaluation of a pharmacist-led digital hypertension management program in a tertiary hospital in Singapore

Supplemental material for Evaluation of a pharmacist-led digital hypertension management program in a tertiary hospital in Singapore by Lee EM, Koh XH, Cheong KM, Khor QH, Chionh CY, Khoo J, Seah A, Ow Yong PE, for the DEMO Study Group in DIGITAL HEALTH.

Supplemental material

Supplemental material - Evaluation of a pharmacist-led digital hypertension management program in a tertiary hospital in Singapore

Supplemental material

Supplemental material - Evaluation of a pharmacist-led digital hypertension management program in a tertiary hospital in Singapore

Footnotes

Acknowledgements

The authors thank Siti Norashikin Binte Fuad (Research Unit) for her assistance in patient recruitment and operational support. Agnes Lim, and Goh Leng Chuan (Pharmacy) for their assistance in patient recruitment. Amilyn Chua (Pharmacy), Yeo Hwee Li and Karen Hay Kai Xin (Office of Innovation) for their administrative support.

ORCID iDs

Elena Meishan Lee

Pu En Ow Yong

Ethical considerations

This study has been approved by the SingHealth Centralised Institutional Review Board (CIRB ref: 2022/2390), and written informed consent was obtained from all participants.

Consent to participate

Research coordinator who explained the study obtained written informed consent from the participants.

Author contributions

EML, PEOY, KMC conceived the study, researched the literature and were involved in protocol development. EML gained ethical approval. EML, PEOY, KMC and JK, were involved in patient recruitment. EML, KMC and PEOY were pharmacists providing the DEMO program. AS extracted data, XHK performed statistical analysis for this research project. EML, PEOY, QHK and XHK wrote the manuscript; All authors reviewed and edited the manuscript and approved the final version of the manuscript.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Joint Research-Innovation Grant (JRIG) (Ref No: RIG202110-014PR). This publication was supported by the Joint Research-Innovation Grant (JRIG) Publication Fund (Ref No: RIG2026-004PF) and CGH Office of Innovation.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

All data generated or analyzed during this study are included in this published article.*

AI disclosure statement

The author(s) used Microsoft Copilot to improve spelling, grammar, clarity, and readability during manuscript preparation. After using this tool, the author(s) carefully reviewed and edited the text and take(s) full responsibility for the final content. The conception, data processing, analysis and content generation of this research manuscript were all independently completed by the authors, without using any artificial intelligence (AI) tools.

Supplemental material

Supplemental material for this article is available online.

References

World Health Organisation . Hypertension 2023. https://www.who.int/news-room/fact-sheets/detail/hypertension (accessed 28 April 2025).

Epidemiology & Disease Control Division, Ministry of Health and Health Promotion Board, Singapore. National Population Health Survey 2022.

Ministry of Health, Singapore . National telemedicine guidelines Singapore, 2015.

Farias

FAC

Dagostini

Bicca

, et al. Remote Patient Monitoring: A Systematic Review. Telemed J E Health 2020; 26(5): 576–583. https://doi.org/10.1089/tmj.2019.0066

Vudathaneni

VKP

Lanke

Mudaliyar

, et al. The Impact of Telemedicine and Remote Patient Monitoring on Healthcare Delivery: A Comprehensive Evaluation. Cureus 2024; 16(3): e55534. https://doi.org/10.7759/cureus.55534

Jackson

Sreedhara

Bostic

, et al. Telehealth Use to Address Cardiovascular Disease and Hypertension in the United States: A Systematic Review and Meta-Analysis, 2011-2021. Telemed Rep 2023; 4(1): 67–86. https://doi.org/10.1089/tmr.2023.0011

Teo

Burkill

, et al. Effects of technology-enabled blood pressure monitoring in primary care: A quasi-experimental trial. J Telemed Telecare 2024; 30(1): 121–130. https://doi.org/10.1177/1357633X211031780

Tan

Kng

Lim

, et al. Long-term Clinical and Cost Outcomes of a Pharmacist-managed Risk Factor Management Clinic in Singapore: An Observational Study. Clin Ther 2017; 39(12): 2355–2365. https://doi.org/10.1016/j.clinthera.2017.10.009

Baral

Volgman

Seri

, et al. Adding Pharmacist-Led Home Blood Pressure Telemonitoring to Usual Care for Blood Pressure Control: A Systematic Review and Meta-Analysis. Am J Cardiol 2023; 203: 161–168. https://doi.org/10.1016/j.amjcard.2023.06.109, Epub 2023 Jul 25. PMID: 37499595.

10.

Chen

Yeh

, et al. Effectiveness of a nationwide community pharmacist-led program promoting home blood pressure monitoring on hypertension control and self-management. Hypertens Res 2026; 49(2): 315–327. https://doi.org/10.1038/s41440-025-02422-6, Epub 2025 Oct 23. PMID: 41131189; PMCID: PMC12823383.

11.

Margolis

Asche

Dehmer

, et al. Long-term Outcomes of the Effects of Home Blood Pressure Telemonitoring and Pharmacist Management on Blood Pressure Among Adults With Uncontrolled Hypertension: Follow-up of a Cluster Randomized Clinical Trial. JAMA Netw Open 2018; 1(5): e181617. https://doi.org/10.1001/jamanetworkopen.2018.1617

12.

Whelton

Carey

Aronow

, et al. ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Hypertension 2018; 71(6): 1269–1324. https://doi.org/10.1161/HYP.0000000000000066 (2017).

13.

Sample size reference Julious

Campbell

. Tutorial in biostatistics: sample sizes for parallel group clinical trials with binary data. Statistics in medicine 2012; 31(24): 2904–2936. https://doi.org/10.1002/sim.5381

14.

Austin

. Optimal caliper widths for propensity-score matching when estimating differences in means and differences in proportions in observational studies. Pharm Stat 2011; 10(2): 150–161. https://doi.org/10.1002/pst.433

15.

Lewis

. The system usability scale: Past, present, and future. International Journal of Human–Computer Interaction 2018; 34(7): 577–590. https://doi.org/10.1080/10447318.2018.1455307

16.

Hofstetter

Miller

Krohmer

, et al. A de-biased direct question approach to measuring consumers' willingness to pay. International journal of research in marketing 2021; 38(1): 70–84. https://doi.org/10.1016/j.ijresmar.2020.04.006

17.

Margolis

Asche

Bergdall

, et al. Effect of home blood pressure telemonitoring and pharmacist management on blood pressure control: a cluster randomized clinical trial. JAMA 2013; 310(1): 46–56. https://doi.org/10.1001/jama.2013.6549

18.

Duan

Xie

Dong

, et al. Effectiveness of home blood pressure telemonitoring: a systematic review and meta-analysis of randomised controlled studies. J Hum Hypertens 2017; 31(7): 427–437. https://doi.org/10.1038/jhh.2016.99

19.

Agarwal

Bills

Hecht

, et al. Role of home blood pressure monitoring in overcoming therapeutic inertia and improving hypertension control: a systematic review and meta-analysis. Hypertension 2011; 57(1): 29–38. https://doi.org/10.1161/HYPERTENSIONAHA.110.160911

20.

Walpola

Issakhany

Gisev

, et al. The accessibility of pharmacist prescribing and impacts on medicines access: A systematic review. Res Social Adm Pharm 2024; 20(5): 475–486. https://doi.org/10.1016/j.sapharm.2024.01.006

21.

Sherman

Pelter

Deamer

, et al. Association between encounter frequency and time to blood pressure control among patients with newly diagnosed hypertension: a retrospective cohort study. J Clin Hypertens (Greenwich) 2018; 20(3): 429–437. https://doi.org/10.1111/jch.13223

22.

Elnaem

Rosley

NFF

Alhifany

, et al. Impact of Pharmacist-Led Interventions on Medication Adherence and Clinical Outcomes in Patients with Hypertension and Hyperlipidemia: A Scoping Review of Published Literature. J Multidiscip Healthc 2020; 13: 635–645. https://doi.org/10.2147/JMDH.S257273

23.

Smith-Ray

Feng

Singh

, et al. Pharmacists as clinical care partners: How a pharmacist-led intervention is associated with improved medication adherence in older adults with common chronic conditions. J Manag Care Spec Pharm 2024; 30(4): 345–351. https://doi.org/10.18553/jmcp.2024.30.4.345

24.

Hong

Wang

Tak

. A patient-centric goal in time to blood pressure control from drug therapy initiation. Clin Transl Sci 2013; 6(1): 7–12. https://doi.org/10.1111/cts.12021

25.

Ang

IYH

Wang

Tyagi

, et al. Preferences and willingness-to-pay for a blood pressure telemonitoring program using a discrete choice experiment. NPJ Digit Med 2023; 6(1): 176. https://doi.org/10.1038/s41746-023-00919-3

26.

Agency for Care Effectiveness (ACE) . Hypertension – tailoring the management plan to optimise blood pressure control. https://www.ace-hta.gov.sg/healthcare-professionals/ace-clinical-guidances/hypertension-tailoring-the-management-plan-to-optimise-blood-pressure-control (2023), (accessed 28 April 2025).

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.88 MB

1.55 MB

0.43 MB

0.00 MB